Modelling of Ground Subsidence from a Combination of GPS and Levelling Surveys

Abstract

Since 1984, the authors have been involved in ground deformation studies in oil fields along the east coast of Lake Maracaibo in Venezuela. Due to oil extraction, an area of about 50 km × 50 km has been subjected to subsidence which in some places reaches 20 cm/year. Subsidence in the whole area has been monitored since 1929 using conventional geodetic levelling with over 1600 benchmarks connected to some points considered to be outside of the subsidence influence. The levelling network consists of 800 km of primary (main network) and 600 km of second-order densification surveys. The complete survey has been repeated every two years with a portion of the network (about a third of the whole area) remeasured every six months for the purposes of upgrading protective coastal dykes, updating the irrigation system, and controlling the stability of offshore platforms and plants. An evaluation (Leal, 1989) of the last few survey campaigns indicates that the standard deviation of the levelling surveys is about 2 mm/√ km for the main network and 4 mm/√km for the densification survey. The average accuracy of the subsidence determination is about 10 mm at a one sigma level. The levelling survey is a slow, expensive, and labour intensive operation. Several survey crews need about three months to measure the whole subsidence area at a total cost of about US$200 000 per campaign.